rstrudel/segmenter
A Vision Transformer-based model for semantic image segmentation released by INRIA researchers.
Not currently ranked — collecting fresh signals.
star history
Segmenter is an official PyTorch implementation of a transformer architecture for semantic segmentation tasks. The model uses an improved Vision Transformer backbone trained on the ADE20K dataset and achieves state-of-the-art results with variants ranging from 7M to 27M parameters. The project is integrated into the MMSegmentation library and provides pre-trained checkpoints for download.
Frequently asked
- What is rstrudel/segmenter?
- A Vision Transformer-based model for semantic image segmentation released by INRIA researchers.
- Is segmenter open source?
- Yes — rstrudel/segmenter is open source, released under the MIT license.
- What language is segmenter written in?
- rstrudel/segmenter is primarily written in Python.
- How popular is segmenter?
- rstrudel/segmenter has 906 stars on GitHub.
- Where can I find segmenter?
- rstrudel/segmenter is on GitHub at https://github.com/rstrudel/segmenter.